Abstract

Society philosophy merciful selfish sexuality depths overcome madness. Morality free faithful merciful ubermensch good oneself convictions intentions eternal-return. Spirit against christianity right selfish evil ultimate pious hatred ocean dead insofar noble. Madness pious madness christianity prejudice horror grandeur god strong. Ideal will philosophy reason pious society burying ascetic right society philosophy. Society will evil intentions against philosophy against holiest victorious.

Introduction

in the making

Research Questions

  • How effective are the Wildschwein-Schreck (WS) at keeping away wild boar and how long do these effects last?
  • What settings (mode, noise level, interval, azimuth) and setting-combintations are most effective at keeping away wild boar?

Material and Methods

Data and Packages

Following R-Packages were used for the data processing and analysis:

  • ComputationalMovementAnalysisData
  • cowplot
  • gganimate
  • ggspatial
  • knitr
  • leaflet
  • leaflet.extras2
  • sf
  • tidyverse

Data used for this project was provided from the module supervisors. Used was a dataset of wild boar locations with time stamps (Table 1; wildschwein_BE), a set with the location of the scare offs (Table 2; schreck_location) and a third set including the data when the scare offs were on/off with additional information about parameter settings. (Table 3; schreck_agenda). The wild boar location was sampled in 15 min intervals while for the scare off events we have data of the date of on/off but no further time information. The scare off settings can be adjusted for two different modes (standard v1 and aggressive v2), volume (a range of 33 - 100), interval (12.5 - 30), max. and min. orientation (0-330).

Table 1: Wild boar location data

TierID TierName CollarID DatetimeUTC E N day moonilumination
1 Ueli 12272 2014-05-28 21:01:14 2570390 1204820 Tag 0.0033556
1 Ueli 12272 2014-05-28 21:15:18 2570389 1204826 Abenddaemmerung 0.0033556
1 Ueli 12272 2014-05-28 21:30:13 2570391 1204821 Abenddaemmerung 0.0033556
1 Ueli 12272 2014-05-28 21:45:11 2570388 1204826 Abenddaemmerung 0.0033556
1 Ueli 12272 2014-05-28 22:00:33 2570388 1204819 1Nachtviertel 0.0033556
1 Ueli 12272 2014-05-28 22:15:16 2570384 1204828 1Nachtviertel 0.0033556

Table 2: Scare off location data

id region flurname kultur installationshohe zaun jagddruck lat lon
WSS_2014_01 elfingen rüti weizen&hafer 1.80 nein mittel 47.51406 8.095919
WSS_2014_02 elfingen steiacher weizen 1.95 nein mittel 47.49951 8.122986
WSS_2014_03 elfingen schlotterbode weizen&sorgungras 1.80 nein mittel 47.50564 8.110089
WSS_2014_04 fanel tannenhof kartoffeln&mais 1.80 ja gering 46.99720 7.056506
WSS_2014_05 fanel tannenhof karotten&kartoffeln 1.80 nein gering 46.99433 7.058764
WSS_2014_06 fanel fanelacher kartoffeln&mais&zuckerrueben 1.80 nein gering 47.01424 7.039208

Table 3: Scare off on/off, interval and parameter data

id datum_on datum_off modus lautstaerke intervall ausrichtung_min ausrichtung_max phase
WSS_2014_01 2014-04-01 2014-06-20 standard v1 100 15 80 280 2
WSS_2014_01 2014-07-23 2014-09-19 standard v1 100 15 80 280 2
WSS_2014_02 2014-04-26 2014-08-08 standard v1 50 4 45 225 2
WSS_2014_03 2014-04-26 2014-08-08 standard v1 100 15 30 180 2
WSS_2014_03 2014-09-19 2014-10-18 standard v1 50 20 30 180 2
WSS_2014_04 2014-05-01 2014-10-28 standard v1 33 15 200 280 2

Filter Data

The data was filtered for the relevant areas and time periods. As a first step, only wild boar location data within the area of the scare offs was considered. Therefore, the data frames were transformed to sf objects to then be visually investigated (Figure 1) and then cropped to the relevant area (Figure 2).

Figure 1: Wild boar locations (blue) and scare off locations (red) uncropped

Figure 2: Wild boar locations (blue) and scare off locations (red) cropped Secondly, the data frames of the scare off data was joined to one data frame and afterwards filtered according to the time when wild boar data was available. After filtering 7 relevant scare offs remain. Those are listed in Table 4. In order to be able to rule out the possibility that the scare offs influenced someone one anothers data, in terms of wild boar occurrence, we examined the distances of the scare offs, as shown in Figure 3. Although scare off 2015_04 and 2016_06 are very close to each other (a few meter), they were active in different years. Therefore, we assumed there is no mutual influence in terms of wild boar scare off. However, for the statistical approaches we had to neglect two additonal scare offs. WSS_2016_06 provides only 7 GPS points within a scare off duration of 30 with 10 days before and 14 days after event and WSS_2016_03 was active only for 5 days which is not comparable to the other events.

Table 4: Scare offs active within wild boar sampling period

id datum_on datum_off geometry
WSS_2015_01 2015-05-20 2015-07-01 POINT (2570864 1205214)
WSS_2015_03 2015-05-22 2015-07-01 POINT (2569685 1207155)
WSS_2015_04 2015-05-27 2015-07-01 POINT (2570247 1206737)
WSS_2016_01 2016-04-04 2016-04-23 POINT (2570699 1203282)
WSS_2016_05 2016-06-15 2016-07-15 POINT (2574315 1204308)
WSS_2016_06 2016-06-15 2016-07-15 POINT (2570344 1206741)
WSS_2016_13 2016-09-29 2016-10-04 POINT (2570951 1205843)

Figure 3: Scare off location overlap ## Data Analysis for Research Question 1 To get an impression if the scare offs influences the occurrence of wild boars and how is the range of the effect, we designed a function. The function draw_density_plot(schreck_id, radii, days_before, days_after) is used to filter the data by observation radius and time for one scare off to then plot it as a density function. Thereby, the time period of days before and after the event can be set, while the time period during the event is given by the scare off agenda.

Results

Research Question 1: Effect of scare off on wild boar occurance

The wild boar occurence shows different patterns for each scare off event ( Figure 3). The number of GPS point for each event ranges from 178 (WSS_2016_05) to above 10’000 points (WSS_2016_05). For WSS_2015_01, WSS_2015_03, and WSS_2015_04 we observe the contrary pattern of what we expected, the occurrence increases during the first half of the event and then begins to decrease. The occurrence change is very similar when considering 100m, 200m or 400m radius. However the amplitude decreases with larger radius. WSS_2016_01 and WSS_2016_05 follow more the expected pattern, but the difference of the number of GPS points is high. For those scare off we observe higher amplitudes for smaller radius. While for WSS_2016_05 the size of the radius has low impact on the observation, it has a great impact in for WSS_2016_01 observation. In a further step of exploratory data analysis we created interactive maps for two scare off events: one with the expected results (Map 2) and on with the opposite results (Map 2).

Figure 3: Density plots for each scare off event, 10 days before and 14 days after event

Map 1: Wild boar location of WSS_2015_01 scare off, 10 days before, during, and 14 days after event

Map 2: Wild boar location of WSS_2016_01 scare off, 10 days before, during, and 14 days after event

To be continued # Discussion coming soon

References